| Summary |
In recent years, international aviation organizations and related industries have paid more attention to the rights of air passengers. At the same time, more and more people choose to use passenger complaints to express individual or group dissatisfaction. Due to data availability of air service passengers' complaints, the literature related to air service passengers' complaints is extremely limited. Even if there are research that mention customer complaints, few discussed in a quantitative analysis method. It is hard to provide effective and objective evidence for the parts of airline service failures that need to be improved. This research estimates econometric models for international and domestic routes respectively, to quantify the factors affecting the number of complaints.
Since the number of complaints on international routes has a relatively large percentage of zero and the data type is over-dispersed, we suggest to use the Zero-Inflated Negative Binomial Model for this market. Estimation results indicates that the flight execution rate, the proportion of server, the number of passengers, the average flight distance of airlines, and the occurrence of strikes have significant impacts on the number of complaints on international routes. Two Negative Binomial Models with and without considering Zero Inflation effects give similar qualitative results, although the coefficients may have some differences. However, the OLS Model leads to biased results. For domestic routes, because each sample comes from a combination of different airlines and routes, we recommend to use Cross-Classified Multilevel Model for this market. Estimation results show that numbers of denied boarding and abnormal baggage, load factor, route market concentration, and under the condition of monopoly market have significant impacts on the number of complaints. In addition, since average flight distance is an uncontrollable factor affecting complaints for international routes, the civil aviation authority might consider adjusting the service evaluation standards according to this factor, in order to improve the fairness of the service evaluation |
